Monthly Archives: February 2006

Future of GNOME = future of free desktops !?

John Williams asks about the future of GNOME. I would agree with his analysis, though not with his consequences. I think the problem is the definition of GNOME itself. Historically it was important to build GNOME as an organisation as well as a desktop. Companies have since built uppon GNOME. The problem today is, that the software world is changing quickly. Since the desktop environments KDE and GNOME have hit the market they gained at least some popularity in the Linux community an beyond. But the demands of the users on free software is growing. We have solved many of the basic issues. Just to remind you we now have (more than one) free Office Suite with OpenOffice.org, we have competing free software browsers that have even got to be implemented in Safari of MacOS as the standard browser. Both major DEs are much more stable and powerfull today. The range of users, that are using Linux and free desktops has been growing massively. The future is bright, especially if you think about Asia, Africa ans South America. Linux will quickly spread everywhere, where money is essential. And if you think of special business software (where only one application will run on one computer): It will not matter what operating system it is installed on – so companies could use Qt odr Gtk as a basis and some kind of kiosk mode.
So what are we missing? I think the greatest lack we have is still collaboration. Bug comanies like Microsoft, Sun, Google, Oracle or Salesforce.com have a high organisational grade. They can easily change direction and it is more easy to organize ressources for a common goal. The freedom in free software allows us to change quick in many different small projects. There is no ONE Linux or ONE KDE direction. Many projects follow their own goals, What is essential and should be thriving is what we have in common.

But people tend to think in groups. This makes things easier if you want to make small decisions, but it also makes things complicated if the groups are not open to each other. And we still have that with still rivaling communities like those of KDE and GNOME. In 2005 the core goals are the same, as are to build a desktop based on free software. jeff Waugh declared an ambitious goal in the past, that was to have 10 percent of the desktop market fpr GNOME in 10 years. I am completely convinced that we will not even touch this goal without giving up the strict community thinking. We have a lot of great communities. Just think of: WINE , Apache, Drupal, WordPress, Wikipedia, Fedora, Ubuntu, GNOME and KDE, linuxprinting.org, GIMP, Inkscape, Mozillas and many many more. Many people in different project know each other and talk and cooperate, but still we do not work together on common goals.

Just one example: In 2005 it is still the situation that if you have GNOME and WINE installed you can install man applications. But there is still no easy way to open a virtual hard disk in GNOMEs Nautilus file manager (the same with konqueror). That means. People might even have a working combination of WINE, a windows application and GNOME – and they download the windows installer and install their application correctly. But the general user who does not know how to open invisible folders will not be able to open the virtual hard drive to work on the files. It seems that nobody yet has thought about that. the problem can be generalized by the word “integration”. You will only find those “bugs” if you think not as a developer only for your own project but begin to think from the users perspective. That’s why commercial projects like Linspire are more successful in targeting users. They are not really better technically, often worse, but they try to solve the users problems. Microsoft also has ever done this. We need to comine two things

  1. make progressive free software
  2. think from the users perspective

I think essentially free software projects have understand No. 1. They have invented this, nobody must tell them. But about No.2: There is a LOOOONNNG way to go. Free software is made by developers. There are marketing experts in the community but most of them sit in professional companies like Red Hat, Sun or Novell and they use the community for the goals of THEIR companies. We should stop to be thankful of what they give us! They have their business model, they feed many of our developers – this is a fair deal. but the free software and free desktop community itself should stand together. I see this as THE essential of free software: People that wanted (as developer and users) their freedom. The freedom will not come by itself.

We as the free software community must communicate to the users. We ARE the users – We ARE the developers – We have even other professionals like journalists, marketing experts (like John Williams) – we even have one millionaire (Mark Shuttleworth ­čśë ). There does not have to be a fight between companies and the community, but it will be better for the companies to follow the community. The current situation where free desktops are still not getting real market share comes from too many companies like Sun that see themselves first, than the customers and on third place the community. I think the community for every free software company must be on the first place and should include their own customers.
And: Yes, i see projects like Wikipedia and even Creative Commons as part of our culture. Software is getting functions to put free licences in graphics. Free content and free software combined are much more powerful. They are even more powerful if we add free standards like from Xiph.org (OGG/Theora), Jpeg, PNG, XML to this soup.

3 Comments

Filed under Free Culture, Free Software

A network window manager

Just as I browsed the Wikipedia about window managers I stumpled uppon WinLIKE, a window manager programmed with Javascript/Ajax. Then I got the idea of a new kind of window manager:

  • imagine the window manager is only installed on a server and announcing itself via Bonjour (uses HTTP or not?)
  • If the Client starts up he gets the login of the window manager from Bonjour. I know we now have XDMCP but I think this is not appropriate, today
  • The window manager should be very lightweight and give the user a configurable interface. I mean it should have a Kiosk mode as a default – so maybe normally you start with a browser or whatever the admin gives you.
  • The configuration and administration should be very easy, similar to Sabayon. It should also allow the administrator to switch the interface to what the user needs – from a simple juke box, to a kiosk browser, an accounting system, just displaying movies. The admin also should be able to change the display in real time without having to use VNC or similar.
  • It should also be possible to share or switch screens if people switch their places where they work, even switching devices,
  • it should also be possible to “freeze” a session and continue work at exactly the same point where it had left. Well this would depend highly on virtualisation on the side of the OS, doesn’t it?
  • I suppose all the techniques should be more server based (editors, email,…) the client only gets very small applications with only the very necessary size, wherefore the real power is on the server.
  • It should let users easily share and cooperate (like with Wikis), also conferencing should be implemented.
  • Most of the techniques are set up to be used internally, although potentially this should also work if you combine two servers and there should be gateways to the outside for may, conferencing,… .

This all is nothing new, think about network computer XDMCP, network boot, Ajax. But right now I do not have a system where I have everything┬á like this on a server. Most window managers do very much depend on local interaction, especially when it comes to sound – and what do you do with USB devices? What I would envision is a system that is lightweight on the one hand and powerful in its integrated applications on the other hand. A sort of bringing back webapplications to an intranet whereas they are not just web pages. A web page is the clients desktop and there is no browser. I don’t think this should be made with Ajax, i don’t think Ajax is powerful enough. Ajax is nice for web access where everything is already slow and users are used to wait.

A lightweight window manager, that also has no problems with different devices like PDAs,…

Comments welcome.

Thilo

Leave a comment

Filed under Uncategorized